In recent years, a processing speed of a graphic processing unit (GPU) becomes important to allow a display to display a high-definition three-dimensional image.
A general GPU divides a display area into multiple tiles and extracts a pixel that exists in an area in which each of the tiles and an area of a primitive such as a triangle overlap each other. Then, the GPU allocates the extracted pixel to any one of multiple parallel processing groups. Further, the GPU performs processing such as texture mapping or bump mapping on each group in parallel and calculates a color of each pixel.
Patent Document 1 discloses that a screen on which the primitive is displayed is divided into multiple blocks and each of the multiple blocks is associated with a bin. Further, Patent Document 1 discloses that the GPU detects the bin overlapping the primitive and allows a unified shader system to process a pixel belonging to the bin.